AmAUJJast-Sm-}  ^  Vanillin  Tests.  393 
tionship,  is  usually  borne  in  mind  by  the  careful  worker  in  this  field 
of  chemistry,  and  it  is  customary  to  apply  all  of  the  known  tests 
for  the  identification  of  an  organic  body  before  deciding  definitely 
regarding  it. 
A  sample  of  ice-cream  was  recently  brought  to  the  author  of  this 
paper  for  examination,  with  the  information  that  it  had  been  reported 
to  contain  formaldehyde.  An  examination  by  three  well-known 
methods  apparently  indicated  the  unmistakable  presence  of  formal- 
dehyde. These  methods  were  the  Hehner  contact  test,  the  resorci- 
nol-sulphuric  acid  contact  test  and  the  phenol-sulphuric  acid  contact 
test.  As  the  author  is  in  the  habit  of  always  applying  the  phenyl- 
hydrazine  test,  this  test  was  applied  with  negative  results,  and  a 
further  application  of  the  phloroglucol,  resorcinol-soda,  and  hydro- 
chloric  acid  tests  also  gave  negative  results.  The  flavor  of  the 
ice-cream  was  easily  recognized  as  vanilla,  and  vanillin  being  an 
aldehyde,  and  thus  indirectly  related  to  formaldehyde,  it  was  con- 
sidered advisable  to  make  some  experiments  with  this  substance 
with  a  view  to  ascertaining  whether  it  gave  similar  reactions  to 
formaldehyde  with  those  tests  which  had  indicated  the  presence  of 
that  substance. 
A  solution  of  vanillin,  101Q1),  was  made  up  and  employed  in  the 
various  tests,  and  it  was  found  to  produce  color  reactions  in  ail  of 
the  zone  tests,  which  were  either  identical  in  appearance  with  the 
colors  produced  by  known  solutions  of  formaldehyde  which  were 
tested  at  the  same  time,  or  were  so  close  a  resemblance  as  to  render 
comparison  necessary  in  order  to  distinguish  them. 
A  sample  of  milk  was  then  flavored  with  the  vanillin  and  distilled 
and  the  reactions  applied  to  the  distillate  with  similar  results. 
Further  investigation  of  the  subject  showed  that  artificial  vanillin 
and  the  vanillin  contained  in  an  extract  made  from  the  genuine 
vanilla  bean  behaved  in  exactly  the  same  manner,  and  that  unless 
the  phenylhydrazine,  phloroglucol,  or  one  of  the  other  tests  men. 
tioned  as  not  producing  the  reaction,  were  applied,  the  presence  of 
formaldehyde  in  the  solution  would  unhesitatingly  be  affirmed. 
It  was  considered  desirable  to  know  in  this  connection  whether 
coumarin,  which  is  sometimes  associated  with  vanillin  in  the  cheaper 
extracts,  would  produce  similar  results,  but  the  results  with  every 
one  of  the  tests  as  applied  to  coumarin  were  entirely  negative. 
The  following  table  of  experiments  upon  a  number  of  the  sub- 
